CharGen: So You Think Can Dance With A Purist?

Back in May 2009, Shamus Young, over at Twenty Sided, wrote a blog regarding the distribution curves that are generated as a result of different character creation methods.

The table you see, above (from Shamus' post) is the distribution of the die-roll results based on the traditional (purist) 3d6 method of character generation. As you can see, the results of a 3d6 character generation method are heavily skewed to the 7-14 score range, with the mean (average) score being 10.5. Most people that have been playing RPGs for some time, are well aquainted with distribution curves, as we tend to pay attention to the related probabilities in our metagaming.

I got a chuckle out of Shamus' opening paragraph to his blog post:

"Based on the comments in the previous post, it seems like many players generate their characters using the following method:
1. Roll 4d6
2. Discard the lowest number
3. Add the remaining three together
4. Wait until the DM isn’t looking
5. Write down whatever numbers you want.
6. Make sure one of them is a 9, just to keep yourself honest."

To avoid this tendency among players, a DM, who is an OD&D purist, might insist that she see every character generation die roll, and that the player record them, in stat-order, for good or ill. But what DM has the time, or inclination, to do this, and where does that leave the trust level between the DM and her players?

In addition, despite what the distribution curve tells us, there are times when someone will have a particularly bad or spectacularly lucky run of dice rolling, and will have either a hopeless character or a super character. Recent versions of D&D have tried to smooth out this variation by implementing the "point-buy" system for character generation, which I think is both a terrible way to create characters, and a contributor to the general diminishment of the art of role-playing.

I promote the 3d6 in stat-order method of character generation. However, I also recognize that players may occasionally be unlucky, lucky, or dodgy in their character generation. My solution, which is definitely NOT purist. is to modify the character stats, after the character is rolled.

If the dice rolls have come up according to the standard distribution, the sum of the scores of all six character stats should be 63 (average score of 10.5 x 6 character stats = 63). If the sum of the character stats are less than 63, I let the player roll the same number of d6's as her character's stats are short of 63.

For example, if the sum of her 6 character stats is 59, she would roll 4d6, one d6 for each point below 63. The number rolled on each d6 corresponds to the stat, in whatever order they are written down. If Strength is in position one, then a die roll of one means strength, if Intelligence is in position two, then a die roll of two means intelligence, and so on. In this example, if she rolled two "1's" and two "2's", she would increase both her Strength and Intelligence by two each.

If the sum of a character's stats are between 63 and 69, I let the player play that character without modification.

However, if the sum of the character's stats exceed 69, I apply the same method (used to improve the below-average character) but in reverse. So again, say the sum of the character's stats are 75, then the player would roll 6d6, since 75 is six higher than 69. The digit rolled on each d6 would correspond the stat holding that position on the character sheet. The player would reduce their score, in each stat, by one point for each time that number was rolled on one of those d6's.

Most people would be okay with a DM allowing the improvement of a "hopeless" character, but why would you penalize a player who got lucky with their dice rolls? The most important reason is to unsure game balance between the players: it's not fun when the characters with average scores are outshone by the super characters, simply because of ability scores. The point of having random dice-roll modifications to the hopeless and super characters, is that it fits with the OD&D philosophy, that you don't pick your stats.

So now you know. When it comes to character generation, I am not a purist, I can house-rule with the best of them.